Re: [libvirt] [PATCH 06/14] Define a wire protocol for talking to the virtlockd daemon
On 12.12.2012 19:32, Eric Blake wrote:
On 12/12/2012 11:14 AM, Michal Privoznik wrote:
> On 11.12.2012 21:41, Daniel P. Berrange wrote:
>> From: "Daniel P. Berrange" <berrange(a)redhat.com>
>>
>> The virtlockd daemon will be responsible for managing locks
>> on virtual machines. Communication will be via the standard
>> RPC infrastructure. This provides the XDR protocol definition
>>
>> * src/locking/lock_protocol.x: Wire protocol for virtlockd
>> * src/Makefile.am: Include lock_protocol.[ch] in virtlockd
>>
>> Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com>
>> ---
>> .gitignore | 1 +
>> cfg.mk | 3 ++
>> src/Makefile.am | 14 ++++++-
>> src/locking/lock_protocol.x | 95 +++++++++++++++++++++++++++++++++++++++++++++
>> 4 files changed, 112 insertions(+), 1 deletion(-)
>> create mode 100644 src/locking/lock_protocol.x
>
> ACK
Should we also tweak src/Makefile.am and add src/lock_protocol-structs
to verify that we don't make any ABI changes to the wire protocol?
We definitely should. However, since this is just an initial
implementation so there's nothing to break so far. Moreover, given how
close to release we are (and how much users wants this in - see Dan's
comment in the cover letter) I think it is not a show stopper for now.
However, if he adds it just before pushing that would be great.
On the other hand - virlockd is to be shipped within libvirtd package.
And it's not to be used by users directly. So the requirements to
backward compatibility seems slightly relaxed to me when compared to
virsh+libvirtd. But then again - that's not a green light to breaking
the compatibility, it's rather the opposite.
Michal